Soft Sensing Model Visualization: Fine-tuning Neural Network from What Model Learned

TL;DR

This paper introduces a deep learning-based soft sensing model for wafer defect detection, utilizing visualization techniques to interpret and fine-tune the model for improved understanding and performance in complex manufacturing data.

Contribution

It presents a novel application of deep visualization for interpreting and guiding the fine-tuning of neural networks in wafer defect detection with imbalanced datasets.

Findings

Effective defect detection on imbalanced data

Visualization aids in understanding model decisions

Fine-tuning improves model interpretability and accuracy

Abstract

The growing availability of the data collected from smart manufacturing is changing the paradigms of production monitoring and control. The increasing complexity and content of the wafer manufacturing process in addition to the time-varying unexpected disturbances and uncertainties, make it infeasible to do the control process with model-based approaches. As a result, data-driven soft-sensing modeling has become more prevalent in wafer process diagnostics. Recently, deep learning has been utilized in soft sensing system with promising performance on highly nonlinear and dynamic time-series data. Despite its successes in soft-sensing systems, however, the underlying logic of the deep learning framework is hard to understand. In this paper, we propose a deep learning-based model for defective wafer detection using a highly imbalanced dataset. To understand how the proposed model works, the deep visualization approach is applied. Additionally, the model is then fine-tuned guided by the deep visualization. Extensive experiments are performed to validate the effectiveness of the proposed system. The results provide an interpretation of how the model works and an instructive fine-tuning method based on the interpretation.